Ocean subsurface temperature inversion method and device and terminal equipment

Abstract

The invention provides an ocean subsurface layer temperature inversion method and device and terminal equipment, and is suitable for the technical field of data processing, and the method comprises the steps: obtaining a plurality of pieces of ocean observation information; according to the multiple pieces of ocean observation information, training an initial sub-surface layer temperature inversion model to obtain multiple target sub-surface layer temperature inversion models; and on the basis of the multiple target subsurface layer temperature inversion models, according to the multiple pieces of ocean observation information and a preset ocean observation depth threshold value, ocean subsurface layer temperature information is calculated. According to the method, the mapping relation between the deep layer temperature and the sea surface information is constructed through the subsurface layer temperature inversion model, the method is used for carrying out inversion calculation on the ocean subsurface layer temperature, and accurate and effective data support is provided for ocean underwater environment research work.

Description

Technical Field

[0001] The present application belongs to the field of data processing technology, and in particular to an ocean subsurface temperature inversion method, device and terminal equipment. Background Art

[0002] As the largest region on Earth, the ocean plays a core role in global climate regulation, energy transmission, and water cycle. Its complex ocean circulation not only exchanges energy and matter with the atmosphere, but also plays a vital regulatory role in the global climate system. Behind these dynamic processes, the temperature distribution of seawater is one of the key driving factors. The ocean temperature field not only determines the density changes of seawater, thereby affecting the flow pattern of the ocean, but also plays an important role in sea-air interaction, ocean circulation, and global climate change. Therefore, accurately estimating the three-dimensional temperature structure below the ocean surface is of great scientific significance for studying these dynamic processes and their impact on the global climate system.

[0003] Traditional observation methods mainly rely on buoys, ships and other platforms for on-site observations. The sea surface information obtained by satellite remote sensing and on-site observations is input into a specific computing network for analysis, thereby calculating the temperature data of the ocean surface.

[0004] However, traditional observation methods obtain underwater temperature data over a large area, and satellite remote sensing technology can only obtain information on the ocean surface, which cannot be used to determine the underwater dynamic environment and its three-dimensional temperature distribution. Most of the global ocean surface observation data obtained through the rapid development of satellite remote sensing technology are based on surface data, ignoring the correlation between temperatures at different depths, and only using surface data to invert the underwater environment field. Therefore, as the depth increases, the inversion accuracy decreases more and more obviously. Summary of the invention

[0005] In view of this, the embodiments of the present application provide a method, apparatus and terminal device for inverting the subsurface temperature of the ocean, aiming to solve the problem in the prior art that it can only calculate the temperature of the surface of the ocean but cannot analyze and solve the subsurface temperature of the ocean, and cannot determine the underwater temperature distribution of the ocean.

[0006] A first aspect of an embodiment of the present application provides an ocean subsurface temperature inversion method, comprising: Obtain multiple ocean observation information; Training an initial subsurface temperature inversion model according to the plurality of ocean observation information to obtain a plurality of target subsurface temperature inversion models; Based on multiple target subsurface temperature inversion models, the ocean subsurface temperature information is calculated according to multiple ocean observation information and a preset ocean observation depth threshold.

[0007] A second aspect of an embodiment of the present application provides an ocean subsurface temperature inversion device, comprising: An ocean observation information acquisition module is used to acquire multiple ocean observation information; A subsurface temperature inversion model training module, used to train an initial subsurface temperature inversion model according to the plurality of ocean observation information to obtain a plurality of target subsurface temperature inversion models; and The ocean subsurface temperature information calculation module is used to calculate the ocean subsurface temperature information based on multiple target subsurface temperature inversion models, multiple ocean observation information and a preset ocean observation depth threshold.

[0008] A third aspect of an embodiment of the present application provides a terminal device, which includes a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements the steps of the ocean subsurface temperature inversion method described in the first aspect above.

[0009] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, comprising: a computer program stored therein, which, when executed by a processor, implements the steps of the ocean subsurface temperature inversion method as described in the first aspect above.

[0010] Compared with the prior art, the embodiments of the present application have the following beneficial effects: the nonlinear relationship between sea surface observation data and underwater temperature is captured through a subsurface temperature inversion model, a mapping relationship between deep temperature and sea surface parameters is constructed, and a connection between ocean temperatures at different depths is established. The subsurface temperature inversion model is trained through the acquired ocean observation data, and then the correlation between ocean temperatures at different depths is used to invert the ocean subsurface temperature field layer by layer, thereby improving the efficiency of inversion calculation of the ocean subsurface temperature. Compared with the inversion of a model without adding the correlation between ocean temperatures at different depths, the mean square error is improved by 5.61%, overcoming the complexity of calculating the ocean underwater temperature field and providing accurate and effective data support for ocean underwater environment research. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0012] Figure 1 This is a schematic diagram of the implementation process of the ocean subsurface temperature inversion method provided in Example 1 of the present application; Figure 2 This is a schematic diagram of the implementation process of the ocean subsurface temperature inversion method provided in Example 2 of the present application; Figure 3 This is a schematic diagram of the implementation process of the ocean subsurface temperature inversion method provided in Example 3 of the present application; Figure 4 This is a schematic diagram of the implementation process of the ocean subsurface temperature inversion method provided in Example 4 of the present application; Figure 5 This is a schematic diagram of the implementation process of the ocean subsurface temperature inversion method provided in Example 5 of the present application; Figure 6 It is a schematic diagram of the structure of the ocean subsurface temperature inversion device provided in an embodiment of the present application; Figure 7 It is a schematic diagram of a terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0013] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0014] In order to illustrate the technical solution described in this application, a specific embodiment is provided below for illustration.

[0015] Figure 1 The following is a flowchart of the implementation of the ocean subsurface temperature inversion method provided in Example 1 of the present application, which is described in detail as follows: Step S101, obtaining a plurality of ocean observation information.

[0016] In this embodiment, the ocean observation information can be obtained through satellite remote sensing and field observation. It can be understood that the ocean observation information needs to be preprocessed, and the preprocessed data can be used to train the initial subsurface temperature inversion model.

[0017] In this embodiment, preferably, the ocean observation information may include observation time information, observation area latitude and longitude information, sea surface height information, sea surface salinity information, sea surface wind speed information, sea surface temperature information, and multiple subsurface temperature observation information. It can be understood that the area and time selected for the monthly average observation data of sea surface height (SSH), sea surface salinity (SSS), sea surface temperature (SST), sea surface wind (SSW), and subsurface ocean temperature (ST) obtained based on satellite remote sensing and field observations are the same.

[0018] Step S102: training an initial subsurface temperature inversion model based on the plurality of ocean observation information to obtain a plurality of target subsurface temperature inversion models.

[0019] In this embodiment, the initial subsurface temperature inversion model can be a CNN model. The preprocessing process for the ocean observation information can include determining the abnormal value of the ST at a depth of 1000m. If the temperature is less than 0°C or the temperature is greater than 10°C, it is determined to be a temperature anomaly area, and the data in the temperature anomaly area is deleted and filled with nan values. If the boundary area exceeds the 9×9 selected range, the excess part is filled with zeros to indicate the lack of information. The preprocessed data set can be divided into a training set and a validation set. The initial subsurface temperature inversion model after training is the target subsurface temperature inversion model. Through training, it is used to minimize the mean square of the difference between the trained neural network prediction value and the actual observation value under the same conditions to obtain multiple target subsurface temperature inversion models. It can be understood that in this embodiment, a large amount of ocean observation information is divided into several homologous data, which are used to train multiple identical initial subsurface temperature inversion models respectively, so that multiple target subsurface temperature inversion models can be obtained.

[0020] Step S103, based on multiple target subsurface temperature inversion models, multiple ocean observation information and a preset ocean observation depth threshold, calculate ocean subsurface temperature information.

[0021] In this embodiment, the ocean subsurface temperature at different depths may be calculated by inverting the ocean observation information layer by layer using multiple target subsurface temperature inversion models. Specifically, the sea surface data may be first used as input data, and the ocean subsurface temperature at a depth of 5 meters is inverted and calculated through the first target subsurface temperature inversion model, and then the temperature inversion value at a depth of 5 meters is combined with the sea surface data as the input of the second target subsurface temperature inversion model, and then the ocean subsurface temperature at a depth of 20 meters is inverted and calculated, and so on, the ocean subsurface temperatures at 24 depths of 5m, 20m, 40m, 60m, 70m, 80m, 90m, 100m, 125m, 150m, 175m, 200m, 225m, 250m, 275m, 300m, 350m, 400m, 500m, 600m, 700m, 800m, 900m, and 1000m may be inverted and calculated at one time, and the preset ocean observation depth threshold may be manually set, which may be 1000m.

[0022] The ocean subsurface temperature inversion method provided in the embodiment of the present application captures the nonlinear relationship between sea surface observation data and underwater temperature through a subsurface temperature inversion model, constructs a mapping relationship between deep temperature and sea surface parameters, and establishes a connection between ocean temperatures at different depths. The subsurface temperature inversion model is trained through the acquired ocean observation data, and then the correlation between ocean temperatures at different depths is used to invert the ocean subsurface temperature field layer by layer, thereby improving the efficiency of the inversion calculation of the ocean subsurface temperature. Compared with the inversion of a model without adding the correlation between ocean temperatures at different depths, the mean square error is improved by 5.61%, which overcomes the complexity of calculating the ocean underwater temperature field and provides accurate and effective data support for ocean underwater environment research.

[0023] Figure 2 The flowchart of the implementation of the ocean subsurface temperature inversion method provided in the second embodiment of the present application is shown. The difference between the second embodiment and the first embodiment is that the step S103 specifically includes: Step S201, calculating multiple observation parameter variable values ​​according to the sea surface height information, sea surface salinity information, sea surface temperature information, sea surface wind speed information and multiple sub-surface temperature observation information.

[0024] In this embodiment, the calculation formula of the observation parameter intermediate variable value P can be: P=5(SSH+SSS+SST+USSW+VSSW)+q. Among them, SSH is the sea surface height, SSS is the sea surface salinity, SST is the sea surface temperature, SSW is the sea surface wind speed, USSW is the north component of the sea surface wind speed, VSSW is the east component of the sea surface wind speed, and q represents the number of depths. The subsurface temperature observation information can be represented by ST, which can be used to combine with the observation parameter intermediate variable value P to calculate the observation parameter variable value. The combination method can be that the observation parameter intermediate variable value P is used together as the element value of the input tensor composed of the observation parameter variable value, which is used for further inversion calculation of the ocean subsurface temperature.

[0025] Step S202, generating a plurality of ocean observation parameter tensors according to the plurality of observation parameter variable values, observation time information, and observation area longitude and latitude information.

[0026] In this embodiment, the observation parameter variable value P can be calculated from sea surface height information, sea surface salinity information, sea surface temperature information, sea surface wind speed information and multiple subsurface temperature observation information. SSH is sea surface height, SSS is sea surface salinity, SST is sea surface temperature, SSW is sea surface wind speed, USSW is the north component of sea surface wind speed, VSSW is the east component of sea surface wind speed, and q represents the number of depths. The observation time information T can refer to the length of the month, M and N represent the longitude and latitude span of the selected area (for example, if the resolution is 0.25°, it means that the longitude and latitude spans of the selected area are M / 4° and N / 4° respectively), then the generated tensor can be expressed as (T, M×N, L, L, P), where the size of the rectangular box (L, L) is the size of the area where the data point is input, and L is used to represent the horizontal and vertical coordinate values ​​of the rectangular box.

[0027] Step S203: training an initial subsurface temperature inversion model according to the plurality of ocean observation parameter tensors to obtain a plurality of target subsurface temperature inversion models.

[0028] In this embodiment, the input tensor size of the first initial subsurface temperature inversion model can be ((T-12)×M×N,L,L,5), where 5 represents the number of parameters, namely, the five ocean surface parameters of SSH, SSS, SST, USSW and VSSW, and the output tensor size is ((T-12)×M×N,L,L,1), where 1 represents the corresponding ST at the depth. The input tensor size of the second initial subsurface temperature inversion model can be ((T-12)×M×N,L,L,6), where 6 represents the number of parameters, namely, the six ocean surface parameters of SSH, SSS, SST, USSW, VSSW and ST of the first layer, and the output tensor size can be ((T-12)×M×N,L,L,1), where 1 represents the corresponding ST at the depth. And so on, until the initial subsurface temperature inversion model of ST at the last depth is trained, where the input tensor size of the training initial subsurface temperature inversion model is ((T-12)×M×N,L,L,P-1), where 6 represents the number of parameters, namely the P-1 ocean surface parameters of SSH, SSS, SST, USSW, VSSW and ST of the q-1 layers above this layer, and the output tensor size is ((T-12)×M×N,L,L,1), where 1 represents the ST at the last depth, and finally q target subsurface temperature inversion models corresponding to q depths are obtained.

[0029] The ocean subsurface temperature inversion method provided in the embodiment of the present application trains an initial subsurface temperature inversion model through multiple ocean observation information, establishes the correlation between multiple sea surface observation information and ocean subsurface temperature information, and thus inverts and calculates the nonlinearly changing ocean subsurface temperature through sea surface observation data, providing effective data support for the study of changes in the ocean subsurface temperature field.

[0030] Figure 3 The flowchart of the implementation of the ocean subsurface temperature inversion method provided in the third embodiment of the present application is shown. The difference between the third embodiment and the first embodiment is that the step S103 specifically includes: Step S301, calculating the first sub-surface temperature inversion information according to the plurality of ocean observation information and the first target sub-surface temperature inversion model.

[0031] In this embodiment, multiple ocean observation information can be converted into tensor form as the input tensor of the first target subsurface temperature inversion model, and the output result is the first surface temperature inversion information. It can be understood that there are multiple target subsurface temperature inversion models, and the first target subsurface temperature inversion model to be calculated in sequence is used as the first target subsurface temperature inversion model, which is used to calculate the temperature of the shallower subsurface depth, and can be used to invert the temperature at a depth of 5m.

[0032] Step S302, calculating the second surface temperature inversion information according to the plurality of ocean observation information, the first surface temperature inversion information and the second target sub-surface temperature inversion model.

[0033] In this embodiment, the output result of the previous target subsurface temperature inversion model can be combined with the sea surface parameter information to form the input tensor of the second target subsurface temperature inversion model, and the calculation result output by the second target subsurface temperature inversion model is the second surface temperature inversion information. There are multiple target subsurface temperature inversion information, and the second target subsurface temperature inversion model to be calculated in sequence is used as the second target subsurface temperature inversion model, which is used to perform inversion calculation on the output result of the previous target subsurface temperature inversion model combined with the input data composed of the sea surface parameter information, and can be used to calculate the subsurface temperature at a depth of 10m. It can be understood that here only the first target subsurface temperature inversion model and the second target subsurface temperature inversion model are taken as examples, and there can be a third target subsurface temperature inversion model, a fourth target subsurface temperature inversion model, and so on, which are used to perform inversion calculations on the subsurface temperature of the ocean at different depths.

[0034] Step S303, determining whether the ocean depth information corresponding to the second surface temperature inversion information is greater than or equal to a preset ocean observation depth threshold; If yes, the second surface temperature inversion information is determined as ocean subsurface temperature information; If not, generating sub-surface inversion temperature intermediate variable information according to the first surface temperature inversion information and the second surface temperature inversion information; The sub-surface inversion temperature intermediate variable information is used as the first surface temperature inversion information and returned to S302.

[0035] In this embodiment, the preset ocean observation depth threshold may be 1000 meters. It is understandable that the first target subsurface temperature inversion model may be used to calculate the ocean subsurface temperature at a depth of 5 meters, and the second target subsurface temperature inversion model may be used to calculate the ocean subsurface temperature at a depth of 10 meters. After layer-by-layer calculation, the ocean subsurface temperature at a depth of 1000 meters may be finally calculated by the last target subsurface temperature inversion model. Therefore, each time a target subsurface temperature inversion model is inverted, it is necessary to determine whether the depth value corresponding to the calculated ocean subsurface temperature has reached the preset ocean observation depth threshold. If it has reached it, there is no need to perform the next inversion calculation, and the result is directly output as the ocean subsurface temperature information; if it has not reached it, the next inversion calculation is required, that is, the output result of the target subsurface temperature inversion model calculated in this round is combined with the output result of the target subsurface temperature inversion model calculated in all previous inversion rounds and the sea surface parameters to generate a subsurface inversion temperature intermediate variable as the input of the next target subsurface temperature inversion model, so as to calculate the ocean subsurface temperature corresponding to the next depth, so as to realize the inversion calculation of the ocean subsurface temperature at different depths. It can be understood that regardless of whether the depth value corresponding to the calculated ocean subsurface temperature reaches the preset ocean observation depth threshold, the information obtained by these inversion calculations is part of the ocean subsurface temperature information, and can be combined with the ocean subsurface temperature information obtained by the previous inversion calculation and the sea surface observation information to calculate the ocean subsurface temperature information at a greater depth.

[0036] The ocean subsurface temperature inversion method provided in the embodiment of the present application adopts a method of inverting the deep temperature field of the ocean subsurface layer by layer, and inverts the temperatures of different depths of the ocean layer by layer through a target subsurface temperature inversion model. It can be inverted in connection with the temperature characteristics inside the ocean, avoiding the loss of connection information between the internal temperatures of the ocean when directly using characteristics such as sea surface temperature and height, thereby improving the accuracy of the inversion of the ocean subsurface temperature and ensuring the accuracy and effectiveness of the inversion calculation results.

[0037] Figure 4 The flowchart of the implementation of the ocean subsurface temperature inversion method provided in the fourth embodiment of the present application is shown. The difference between the fourth embodiment and the third embodiment is that the step S301 specifically includes: Step S401, generating a plurality of ocean observation numerical tensors according to the plurality of ocean observation information.

[0038] In this embodiment, multiple values ​​in the ocean observation information may be used to generate an input tensor. Specifically, SSH may be sea surface height, SSS may be sea surface salinity, SST may be sea surface temperature, SSW may be sea surface wind speed, USSW may be the north component of the sea surface wind speed, VSSW may be the east component of the sea surface wind speed, q may represent the depth, and the observation time information T, M and N may represent the longitude and latitude span of the selected area (for example, if the resolution is 0.25°, it means that the longitude and latitude spans of the selected area are M / 4° and N / 4° respectively). The generated ocean observation value tensor may be expressed as (T, M×N, L, L, P), wherein the size of the rectangular box (L, L) is used as the area size of the data point input, and L is used to represent the horizontal and vertical coordinate values ​​of the rectangular box.

[0039] Step S402, based on the first target subsurface temperature inversion model, convolution calculation is performed on the multiple ocean observation value tensors to obtain multiple ocean observation value multi-dimensional feature maps.

[0040] In this embodiment, the target subsurface temperature inversion model can be a trained CNN model, and the ocean observation numerical tensor can be convolved through multiple convolutional layers of the CNN model. It can be understood that the CNN model can include two convolutional layers, and the convolutional layer of the CNN model can include a convolution kernel, which can be used to perform convolution calculations with the ocean observation numerical tensor to obtain multiple multi-dimensional feature maps of ocean observation values.

[0041] Step S403, performing dimensionality reduction processing on the plurality of multi-dimensional feature graphs of ocean observation values ​​to obtain a plurality of one-dimensional feature graphs of ocean observation values.

[0042] In this embodiment, the multidimensional feature map of ocean observation values ​​can be reduced in dimensionality through the Flatten layer in the CNN model, and multiple multidimensional feature maps of ocean observation values ​​can be reduced in dimensionality into a one-dimensional matrix to reduce the complexity and amount of subsequent calculations. The matrix obtained after dimensionality reduction is the one-dimensional feature map of ocean observation values.

[0043] Step S404, based on the first target sub-surface temperature inversion model, a nonlinear transformation is performed on the one-dimensional characteristic map of the ocean observation values ​​to calculate the first sub-surface temperature inversion information.

[0044] In this embodiment, the one-dimensional feature map of the ocean observation value can be nonlinearly transformed through the fully connected layer in the CNN model. Specifically, nonlinear mapping and calculation can be performed through the activation function and bias term in the fully connected layer. It can be understood that there can be 3 fully connected layers in the CNN model, and the activation function can be ReLU, which is used to increase the nonlinearity of the subsurface temperature inversion model so that it can handle the complex nonlinear relationship between multiple subsurface depths of the ocean.

[0045] The ocean subsurface temperature inversion method provided in the embodiment of the present application performs convolution calculation on the ocean observation information through the trained subsurface temperature inversion model, extracts features from the ocean observation information, and then performs dimensionality reduction processing on the extracted features to reduce the computational complexity and improve the solution efficiency. The reduced-dimensional features are then mapped to a nonlinear space for transformation, thereby achieving feature combination and conversion, avoiding overfitting of the subsurface temperature inversion model, and improving the efficiency and accuracy of the inversion calculation of the ocean subsurface temperature.

[0046] Figure 5 The flowchart of the implementation of the ocean subsurface temperature inversion method provided in the fifth embodiment of the present application is shown. The difference between the fifth embodiment and the third embodiment is that the step S302 specifically includes: Step S501, merging the plurality of ocean observation information and the first surface temperature inversion information to obtain a sub-surface temperature inversion input tensor.

[0047] In this embodiment, multiple values ​​in the ocean observation information and the first surface temperature inversion information may be combined to generate an input tensor. Specifically, SSH may be the sea surface height, SSS may be the sea surface salinity, SST may be the sea surface temperature, SSW may be the sea surface wind speed, USSW may be the north component of the sea surface wind speed, VSSW may be the east component of the sea surface wind speed, q may represent the number of depths, observation time information T, M and N may represent the longitude and latitude span of the selected area, ST may represent the first surface temperature inversion information, and the generated subsurface temperature inversion input tensor may be represented as (T, M×N, L, L, P, ST), where the size of the rectangular box (L, L) is the area size of the data point input, and L is used to represent the horizontal and vertical coordinate values ​​of the rectangular box.

[0048] Step S502: Based on the second target subsurface temperature inversion model, a convolution calculation is performed on the subsurface temperature inversion input tensor to obtain a plurality of input tensor multi-dimensional feature maps.

[0049] In this embodiment, the second target subsurface temperature inversion model can be a trained CNN model, and the subsurface temperature inversion input tensor can be convolved through multiple convolutional layers of the CNN model. It can be understood that the CNN model can include two convolutional layers, and the convolutional layer of the CNN model can include a convolution kernel, which can be used to perform convolution calculations with the subsurface temperature inversion input tensor to obtain multiple input tensor multi-dimensional feature maps.

[0050] Step S503, performing dimensionality reduction processing on the multiple multi-dimensional feature maps of the input tensors to obtain multiple one-dimensional feature maps of the input tensors.

[0051] In this embodiment, the Flatten layer in the CNN model can be used to reduce the dimensionality of the input tensor multi-dimensional feature map, and multiple input tensor multi-dimensional feature maps can be reduced in dimensionality into a one-dimensional matrix to reduce the complexity and amount of subsequent calculations. The matrix obtained after dimensionality reduction is the one-dimensional feature map of the input tensor.

[0052] Step S504, based on the second target sub-surface temperature inversion model, feature conversion processing is performed on the input tensor one-dimensional feature map to calculate the second sub-surface temperature inversion information.

[0053] In this embodiment, the one-dimensional feature map of the input tensor can be nonlinearly transformed through the fully connected layer in the CNN model. Specifically, nonlinear mapping and calculation can be performed through the activation function and bias term in the fully connected layer. It can be understood that there can be 3 fully connected layers in the CNN model, and the activation function can be ReLU, which is used to increase the nonlinearity of the subsurface temperature inversion model so that it can handle the complex nonlinear relationship between multiple subsurface depths of the ocean.

[0054] The ocean subsurface temperature inversion method provided in the embodiment of the present application performs inversion calculations on ocean observation information and subsurface temperature information through multiple subsurface temperature inversion models, firstly performs feature extraction on the ocean observation information and the output results of the previous subsurface temperature inversion model through convolution operation, and then performs dimensionality reduction processing on the extracted features to reduce the computational complexity and improve the solution efficiency, and then maps the reduced-dimensional features to a nonlinear space for transformation, thereby realizing feature combination and conversion, avoiding overfitting of subsequent subsurface temperature inversion models, and improving the efficiency and accuracy of layer-by-layer inversion calculations of the ocean subsurface temperature.

[0055] Corresponding to the method of the above embodiment, Figure 6 A structural block diagram of an ocean subsurface temperature inversion device provided in an embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown. Figure 6The exemplary ocean subsurface temperature inversion device may be an execution body of the ocean subsurface temperature inversion method provided in the aforementioned first embodiment.

[0056] Reference Figure 6 , the ocean subsurface temperature inversion device includes: The ocean observation information acquisition module 610 is used to acquire a plurality of ocean observation information; A subsurface temperature inversion model training module 620 is used to train an initial subsurface temperature inversion model according to the plurality of ocean observation information to obtain a plurality of target subsurface temperature inversion models; and The ocean subsurface temperature information calculation module 630 is used to calculate the ocean subsurface temperature information based on multiple target subsurface temperature inversion models, multiple ocean observation information and a preset ocean observation depth threshold.

[0057] The process of each module in the ocean subsurface temperature inversion device provided in the embodiment of the present application realizing its own function can be specifically referred to the aforementioned Figure 1 The description of the first embodiment is not repeated here.

[0058] The ocean subsurface temperature information calculation module 630 includes: A first surface temperature inversion information calculation unit, used to calculate the first surface temperature inversion information according to the plurality of ocean observation information and the first target sub-surface temperature inversion model; A second surface temperature inversion information calculation unit, used to calculate the second surface temperature inversion information according to the plurality of ocean observation information, the first surface temperature inversion information and the second target sub-surface temperature inversion model; An ocean sub-surface temperature information determination unit, used to determine whether the ocean depth information corresponding to the second sub-surface temperature inversion information is greater than or equal to a preset ocean observation depth threshold; If yes, the second surface temperature inversion information is determined as ocean subsurface temperature information; If not, generating sub-surface inversion temperature intermediate variable information according to the first surface temperature inversion information and the second surface temperature inversion information; The secondary surface temperature inversion intermediate variable information is used as the first surface temperature inversion information and returned to the second surface temperature inversion information calculation unit.

[0059] The process of each unit in the ocean subsurface temperature information calculation module 630 provided in the embodiment of the present application realizing its respective function can be specifically referred to the aforementioned Figure 3 The description of the third embodiment is not repeated here.

[0060] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0061] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0062] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0063] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0064] In addition, in the description of the present specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish descriptions, and cannot be understood as indicating or suggesting relative importance. It should also be understood that although the terms "first", "second", etc. are used to describe various elements in some embodiments of the present application in the text, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, the first table can be named as the second table, and similarly, the second table can be named as the first table without departing from the scope of the various described embodiments. The first table and the second table are both tables, but they are not the same table.

[0065] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0066] The ocean subsurface temperature inversion method provided in the embodiment of the present application can be applied to terminal devices such as mobile phones, tablet computers, wearable devices, vehicle-mounted devices, augmented reality (AR) / virtual reality (VR) devices, laptops, ultra-mobile personal computers (UMPC), netbooks, personal digital assistants (PDA), etc. The embodiment of the present application does not impose any restrictions on the specific type of the terminal device.

[0067] For example, the terminal device can be a station (STAION, ST) in a WLAN, a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, a Wireless Local Loop (WLL) station, a Personal Digital Assistant (PDA) device, a handheld device with wireless communication function, a computing device or other processing device connected to a wireless modem, a vehicle-mounted device, a vehicle networking terminal, a computer, a laptop computer, a handheld communication device, a handheld computing device, a satellite wireless device, a wireless modem card, a TV set top box (STB), a customer premises equipment (CPE) and / or other devices for communicating on a wireless system and a next-generation communication system, such as a mobile terminal in a 5G network or a mobile terminal in a future evolved Public Land Mobile Network (PLMN) network, etc.

[0068] As an example but not limitation, when the terminal device is a wearable device, the wearable device can also be a general term for wearable devices that are intelligently designed and developed using wearable technology for daily wear, such as glasses, gloves, watches, clothing and shoes. A wearable device is a portable device that is worn directly on the body or integrated into the user's clothes or accessories. Wearable devices are not just hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are fully functional, large in size, and can achieve complete or partial functions without relying on smartphones, such as smart watches or smart glasses, as well as those that only focus on a certain type of application function and need to be used in conjunction with other devices such as smartphones, such as various smart bracelets and smart jewelry for vital sign monitoring.

[0069] Figure 7 Schematic diagram of the structure of a terminal device provided by an embodiment of the present application. Figure 7 As shown, the terminal device 7 of this embodiment includes: at least one processor 70 ( Figure 7 Only one is shown in the figure), a memory 71, in which a computer program 72 that can be run on the processor 70 is stored. When the processor 70 executes the computer program 72, the steps in the above-mentioned embodiments of the ocean subsurface temperature inversion method are implemented, such as Figure 1 Alternatively, when the processor 70 executes the computer program 72, the functions of each module / unit in the above-mentioned device embodiments are realized, for example Figure 6 Functions of modules 610 to 630 are shown.

[0070] The terminal device 7 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The terminal device may include, but is not limited to, a processor 70 and a memory 71. Those skilled in the art will appreciate that Figure 7 It is only an example of the terminal device 7 and does not constitute a limitation on the terminal device 7. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal device may also include an input sending device, a network access device, a bus, etc.

[0071] The processor 70 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0072] In some embodiments, the memory 71 may be an internal storage unit of the terminal device 7, such as a hard disk or memory of the terminal device 7. The memory 71 may also be an external storage device of the terminal device 7, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal device 7. Further, the memory 71 may include both an internal storage unit and an external storage device of the terminal device 7. The memory 71 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program. The memory 71 may also be used to temporarily store data that has been sent or is to be sent.

[0073] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0074] An embodiment of the present application also provides a terminal device, which includes at least one memory, at least one processor, and a computer program stored in the at least one memory and executable on the at least one processor, wherein when the processor executes the computer program, the terminal device implements the steps in any of the above-mentioned method embodiments.

[0075] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0076] An embodiment of the present application provides a computer program product. When the computer program product runs on a terminal device, the terminal device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0077] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0078] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0079] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0080] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0081] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for inverting ocean subsurface temperature, characterized in that: include: Obtain multiple ocean observation information; Training an initial subsurface temperature inversion model according to the plurality of ocean observation information to obtain a plurality of target subsurface temperature inversion models; Based on multiple target subsurface temperature inversion models, ocean subsurface temperature information is calculated according to multiple ocean observation information and a preset ocean observation depth threshold; The ocean observation information includes observation time information, observation area latitude and longitude information, sea surface height information, sea surface salinity information, sea surface wind speed information, sea surface temperature information and multiple sub-surface temperature observation information; The step of calculating ocean subsurface temperature information based on multiple target subsurface temperature inversion models, according to multiple ocean observation information and a preset ocean observation depth threshold, specifically includes: Calculating first subsurface temperature inversion information according to the plurality of ocean observation information and the first target subsurface temperature inversion model; Calculating the second surface temperature inversion information according to the plurality of ocean observation information, the first surface temperature inversion information and the second target sub-surface temperature inversion model; Determine whether the ocean depth information corresponding to the second surface temperature inversion information is greater than or equal to a preset ocean observation depth threshold; If yes, the second surface temperature inversion information is determined as ocean subsurface temperature information; If not, generating sub-surface inversion temperature intermediate variable information according to the first surface temperature inversion information and the second surface temperature inversion information; Using the subsurface temperature inversion intermediate variable information as the first surface temperature inversion information, and returning to the step of obtaining the second surface temperature inversion information according to the plurality of ocean observation information, the first surface temperature inversion information and the second target subsurface temperature inversion model; The step of calculating the first subsurface temperature inversion information according to the plurality of ocean observation information and the first target subsurface temperature inversion model specifically includes: Generating a plurality of ocean observation numerical tensors according to the plurality of ocean observation information; Based on the first target subsurface temperature inversion model, convolution calculation is performed on the plurality of ocean observation numerical tensors to obtain a plurality of multi-dimensional feature graphs of ocean observation numerical values; Performing dimensionality reduction processing on the plurality of multidimensional feature graphs of ocean observation values ​​to obtain a plurality of one-dimensional feature graphs of ocean observation values; Based on the first target sub-surface temperature inversion model, a nonlinear transformation is performed on the one-dimensional characteristic map of the ocean observation value to calculate the first sub-surface temperature inversion information; The step of calculating the second surface temperature inversion information according to the plurality of ocean observation information, the first surface temperature inversion information and the second target sub-surface temperature inversion model specifically includes: Merging the plurality of ocean observation information and the first surface temperature inversion information to obtain a sub-surface temperature inversion input tensor; Based on the second target subsurface temperature inversion model, convolution calculation is performed on the subsurface temperature inversion input tensor to obtain a plurality of input tensor multi-dimensional feature maps; Performing dimensionality reduction processing on the plurality of multi-dimensional feature maps of the input tensors to obtain a plurality of one-dimensional feature maps of the input tensors; Based on the second target sub-surface temperature inversion model, feature conversion processing is performed on the one-dimensional feature map of the input tensor to calculate the second sub-surface temperature inversion information; The initial subsurface temperature inversion model is a CNN model.

2. The ocean subsurface temperature inversion method according to claim 1, characterized in that: The step of training the initial subsurface temperature inversion model according to the plurality of ocean observation information to obtain a plurality of target subsurface temperature inversion models specifically includes: Calculating multiple observation parameter variable values ​​according to the sea surface height information, sea surface salinity information, sea surface temperature information, sea surface wind speed information and multiple sub-surface temperature observation information; Generate multiple ocean observation parameter tensors according to multiple observation parameter variable values, observation time information, and observation area longitude and latitude information; The initial subsurface temperature inversion model is trained according to the plurality of ocean observation parameter tensors to obtain a plurality of target subsurface temperature inversion models.

3. An ocean subsurface temperature inversion device, characterized in that: include: An ocean observation information acquisition module is used to acquire multiple ocean observation information; A subsurface temperature inversion model training module, used to train an initial subsurface temperature inversion model according to the plurality of ocean observation information to obtain a plurality of target subsurface temperature inversion models; as well as An ocean subsurface temperature information calculation module is used to calculate ocean subsurface temperature information based on multiple target subsurface temperature inversion models, multiple ocean observation information and a preset ocean observation depth threshold; The ocean observation information includes observation time information, observation area latitude and longitude information, sea surface height information, sea surface salinity information, sea surface wind speed information, sea surface temperature information and multiple sub-surface temperature observation information; The ocean subsurface temperature information calculation module includes: A first surface temperature inversion information calculation unit, used to calculate the first surface temperature inversion information according to the plurality of ocean observation information and the first target sub-surface temperature inversion model; A second surface temperature inversion information calculation unit, used to calculate the second surface temperature inversion information according to the plurality of ocean observation information, the first surface temperature inversion information and the second target sub-surface temperature inversion model; An ocean sub-surface temperature information determination unit, used to determine whether the ocean depth information corresponding to the second sub-surface temperature inversion information is greater than or equal to a preset ocean observation depth threshold; If yes, the second surface temperature inversion information is determined as ocean subsurface temperature information; If not, generating sub-surface inversion temperature intermediate variable information according to the first surface temperature inversion information and the second surface temperature inversion information; Using the subsurface temperature inversion intermediate variable information as the first surface temperature inversion information, and returning to the step of obtaining the second surface temperature inversion information according to the plurality of ocean observation information, the first surface temperature inversion information and the second target subsurface temperature inversion model; The step of calculating the first subsurface temperature inversion information according to the plurality of ocean observation information and the first target subsurface temperature inversion model specifically includes: Generating a plurality of ocean observation numerical tensors according to the plurality of ocean observation information; Based on the first target subsurface temperature inversion model, convolution calculation is performed on the plurality of ocean observation numerical tensors to obtain a plurality of multi-dimensional feature graphs of ocean observation numerical values; Performing dimensionality reduction processing on the plurality of multidimensional feature graphs of ocean observation values ​​to obtain a plurality of one-dimensional feature graphs of ocean observation values; Based on the first target sub-surface temperature inversion model, a nonlinear transformation is performed on the one-dimensional characteristic map of the ocean observation value to calculate the first sub-surface temperature inversion information; The step of calculating the second surface temperature inversion information according to the plurality of ocean observation information, the first surface temperature inversion information and the second target sub-surface temperature inversion model specifically includes: Merging the plurality of ocean observation information and the first surface temperature inversion information to obtain a sub-surface temperature inversion input tensor; Based on the second target subsurface temperature inversion model, convolution calculation is performed on the subsurface temperature inversion input tensor to obtain a plurality of input tensor multi-dimensional feature maps; Performing dimensionality reduction processing on the plurality of multi-dimensional feature maps of the input tensors to obtain a plurality of one-dimensional feature maps of the input tensors; Based on the second target sub-surface temperature inversion model, feature conversion processing is performed on the one-dimensional feature map of the input tensor to calculate the second sub-surface temperature inversion information; The initial subsurface temperature inversion model is a CNN model.

4. A terminal device, characterized in that: The terminal device includes a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the method according to any one of claims 1 to 2 when executing the computer program.

5. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 2 are implemented.

Images